Rift that will Host Important Deposits of Hydrocarbons: A Review of the Petrologic Constraints Recorded by the Mantle Peridotites from Ophiolite Massifs

Rift formation has long been the focus of attention for researchers, and numerous studies have been carried out in order to understand causes and modes of whole lithospheric extension. The process of lithospheric rifting is classically considered to be a product of “active rifting” or “passive rifting”, depending upon which forces are involved at the inception of rifting. Continental rifting is conventionally described as a thinning process of the whole lithosphere, ultimately leading to rupture of the continent, onset of sea-floor spreading and the consequent formation of a mid-oceanic ridge. Rifting is the initial and fundamental process by which the separation of a continent into two tectonic plates takes place. Active rifting or mantle-activated rifting has been classically ascribed to the ascent of a mantle plume impinging upon the base of the lithosphere, with consequent heating and thinning of the lithosphere. Passive rifting has been classically considered the result of horizontal stretching of the continental lithosphere, in which far-field tectonic stresses, generated at the boundaries of the lithospheric plates, result in lithosphere extension. Continental rifts are the sites of significant oil and gas accumulations, such as the Viking Graben and the Gulf of Suez Rift. Thirty percent of giant oil and gas fields are found within such a setting. In 1999 it was estimated that there were 200 billion barrels of recoverable oil reserves hosted in rifts.


Introduction
Piccardo et al. studied from time the passive continental rifting leading to opening of the fossil Jurassic Ligurian Tethys oceanic basin, by investigating the structural ad petrologic features recorded in the mantle peridotites of the Alpine-Apennine orogeny (North-West Italy), that represent the direct exposure in nature of the mantle lithosphere of the basin. Passive lithosphere extension is testified by km-scale extensional shear zones, induced by far field tectonic forces, which thinned the sub-continental lithosphere and caused the passive upwelling of the asthenosphere (the passive a-magmatic rifting stage). After significant adiabatic upwelling, asthenosphere underwent decompression melting and the melts infiltrated through the extending lithospheric mantle by diffuse porous flow, frequently exploiting porosity bands of former shear zones (the passive magmatic rifting stage). Deformation and melt percolation interacted and mutually enhanced, strongly modifying the rheological characteristics of the mantle lithosphere along the axial zone of rifting (forming a weakened/ softened mantle wedge). The passive rifting system changed to splitting and continental break-up, passing to an active rifting system in the case deeper/hotter asthenosphere actively upwelled within the axial zone of weakened mantle lithosphere. Apennine orogeny (North-West Italy), that represent the direct exposure in nature of the mantle lithosphere of the basin. Passive lithosphere extension is testified by km-scale extensional shear zones, induced by far field tectonic forces, which thinned the sub-continental lithosphere and caused the passive upwelling of the asthenosphere (the passive a-magmatic rifting stage). After significant adiabatic upwelling, asthenosphere underwent decompression melting and the melts infiltrated through the extending lithospheric mantle by diffuse porous flow, frequently exploiting porosity bands of former shear zones (the passive magmatic rifting stage). Deformation and melt percolation interacted and mutually enhanced, strongly modifying the rheological characteristics of the mantle lithosphere along the axial zone of rifting (forming a weakened/ softened mantle wedge). The passive rifting system changed to splitting and continental break-up, passing to an active rifting system in the case deeper/hotter asthenosphere actively upwelled within the axial zone of weakened mantle lithosphere. Both processes produce different rifted continental margins that is Active Rifting : bottom-up processes involving mantle upwelling/plumes (chemical thermal anomalies) drive active rifting and the formation of volcanic rifted (continental) margins and associated sub-aerial LIPs or flood basalt provinces (e.g., The process of "melt thermal advection" means "heat advection by magma transport" [1] or "the effect of advection of heat by ascending melt" [2]. The process of thermal erosion (i.e., "tempera-ture increase in the lithosphere induced by the asthenosphere"), is used, for instance, by Viljoen [3] for the Merensky Reef in the western Bushveld Complex, by Müntener et al. [4], Piccardo [5], Bodinier et al. [6] and many others for alpine orogenic and ophiolitic peridotites, by Griffin et al. [7] Zhang et al. [8,9] and Menzies and Xu [10] for the North China Craton, and by O'Reilly and Griffin [11] for continental cratons. The process of Chemical erosion (i.e., "both chemical depletion and enrich-ment") is used, for instance, by Müntener et al. [4], Piccardo [4], Bodinier et al. [6] and many others for orogenic and ophiolitic peridotites, by Zhang et al. [8,9] and Kusky et al. [12] for the North China Craton, by Xu et al. [13] and Coltorti et al. [11] for CapoVerde, and by Rivalenti et al. [14] for Fernando de Noronha, Carlson [15]. The process of asthenospherization (i.e., "temperature increase in the mantle lithosphere up to asthenospheric values"), is used, for instance, by Bodinier and Godard [16], Piccardo [4], and Müntener et al. [17] for Lanzo, by Vauchez and Garrido [18] and Marchesi et al. [19] for Ronda, by Vauchez et al. [20] for Tanzania xenoliths, by Tommasi et al. [21] for Syberian xenoliths, and by Bascou et al. [22] for Kerguelen, Tang et al. [23] for NE China Craton. The process of Rejuvenation (i.e., "major, trace element and isotopic equilibration of older mantle litho-sphere with younger percolating melts", is used, among others, by Lenoir et al. [24] for Massif Central, by Houseman and Molnar [25] for orogens, by Li et al. [26] for the Hawaiian plume, by Beccaluva et al. [27] for the Hoggar swell, by Foley [28] for continental cratons, by Lister et al. [29] for collision and subduction, by Lorand et al. [30] for the Oman ophiolites, and by Guarnieri et al. [31] for the Lanzo peridotites). Foley [28], Griffin et al. [32], Roy et al. [33] and O'Reilly and Griffin [34] sustained that the sub-continental lithospheric mantle plays an important role in destabilizing continents and tectonic plates, by thermal and chemical modifications during infiltration of melts into the lithospheric mantle column. According to these authors, the meltinfiltration process represents the primary mechanism for weakening and rejuvenating the continental plate through thermal effects. Craton, and by O'Reilly and Griffin [35] for continental cratons. The process of Chemical erosion (i.e., "both chemical depletion and enrichment") is used, for instance, by Müntener [36]. The process of asthenospherization (i.e., "temperature increase in the mantle lithosphere up to asthenospheric values"), is used, for instance, by Bodinier  Griffin sustained that the sub-continental lithospheric mantle plays an important role in destabilizing continents and tectonic plates, by thermal and chemical modifications during infiltration of melts into the lithospheric mantle column. According to these authors, the meltinfiltration process represents the primary mechanism for weak-ening and rejuvenating the continental plate through thermal effects.

Subsolidus Recrystallization, Partial Melting and Melt Percolation
We frequently refer to the subsolidus (metamorphic) transition between spinelto plagioclase-peridotite facies conditions, to melt percolation and interaction under subsolidus conditions, when discussing mantle lithosphere exhumation and asthenospheric melt porous flow migration. In these cases, we do not refer to mantle partial melting at solidus or supra-solidus conditions. Available studies on subsolidusequilibrated plagioclase peridotites have documented that the spinelto plagioclase lherzolite subsolidus transition is a continuous reaction marked by progressive chemical changes in pyroxenes and spinel, in response to the crystallization of plagioclase. This is poorly supported by parallel experimental investigations, be-cause a few studies have been so far performed at subsolidus conditions and in complex chemical compositions approaching those of natural peridotites. Most of the experimental works have been performed at solidus conditions (T N 1200°C), because they addressed mantle partial melting and low-P mid-ocean ridge basalt (MORB) generation [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45]. These studies indicate that the plagioclase-out boundary occurs at pressures between 1.0 and 1.5 GPa, and at solidus temperatures, in variably depleted perido-tites, suggesting a bulk composition dependence [45,46]. Concerning peridotite melting, when mantle peridotite reaches its solidus P-T conditions, it undergoes partial melting and the composition of the formed melt is in equilibrium with the mineralogical assemblage of peridotite. As evidenced by Chalot-Prat et al. [47], long time experimental results give the composi-tions of melt in equilibrium with a mineralogical assemblage at given P-T conditions. In the case of melt infiltration and porous flow percolation through the colder lithospheric mantle, the percolating melts derive from deeper and hotter asthenospheric levels. These melts are, accordingly, not in equilibrium with the percolated mantle lithosphere, and undergo melt/peridotite interaction during reactive porous flow percolation. In fact, mantle lithosphere is at temperature conditions lower than its solidus temperature and it does not undergo a partial melting process but, on the contrary, a melt/rock reaction process (i.e., melt and host peridotite assemblages are not in equilibrium). These processes are well documented by melt/peridotite reaction micro-textural features (clinopyroxene + olivine dissolution, orthopyroxene + plagioclase crystallization) and by compositional characteristics, as they are amply described and documented from decades in the international scientific bibliography. The previous papers [48][49][50][51][52][53][54] evidenced that mantle minerals of the melt-reacted peridotites always show compositional zoning of major and trace elements, suggesting that melt/mineral equilibration of the bulk system was never significantly developed. In rare cases, min-erals show poorly developed zoning and could be interpreted as indicating of approach to melt/mineral equilibration. As a whole, the rocks we studied cannot be considered to represent an equilibrated bulk compo-sition at a particular P and T, that has no mineralogical memory of its origin, since the component minerals preserve clear major, trace and isotopic signatures that unravel their origin and some steps of their previous evolution. In our cases, infiltrated peridotites were heated by melt thermal advection up to 1300°C and some important trace elements were redistributed under equilibrium conditions at those thermal conditions. Further cooling stages did not affect their equilibrium distribution, since diffusion of these trace elements was prevented at lowered thermal conditions. In conclusion, I think that mineral/mineral or mineral/melt equili-bration cannot be simply and solely evaluated on the basis of the distri-bution of the most common major elements, and that trace element and isotope distribution are important tools to unravel the origin and previous histories of minerals and rocks. Chalot-Prat et al. attempting to investigate the origin and re-lationships of processes of magmatism and lithosphere refertilization (i.e., formation of plagioclase peridotites), performed an experimental approach where they have determined the compositions of liquids and co-existing minerals in the six phase assemblage [liquid + olivine + orthopyroxene + clinopyroxene + plagioclase + spinel] at 0.5 GPa and 1100°C to 1200°C. They compared their experimental assem-blages, at 0.5, 0.75 and 1 GPa, with the compositions of minerals from plagioclase ± spinel lherzolite at Lanzo. Field, petrological and geochemical studies evidenced that the Lanzo plagioclase peridotites are 'refertilized' (plagioclase + orthopyroxene crystallization, olivine + clinopyroxene dissolution) by the reaction of residual peridotite with percolating silica-saturated derivative basaltic magma. Accordingly, Chalot-Prat et al. applied the results of experimental equilibrium partial melting to natural melt/peridotite disequilibrium interaction and impregnation. Chalot-Prat et al. concluded that Lanzo plagioclase ± spinel lherzolites equilibrated at pressures between 0.75 and 1 GPa, at temperatures ~ 100-200°C below the solidus and their experimental study suggests that the process of refertilization took place at depths of 25-30 km. In any case, these assertions confirm previous P-T estimates for the formation of the South Lanzo plagioclase peridotites. It must be recalled, as summarized by Piccardo (in press, and references therein) on the basis of knowledge on the Ligurian Tethys plagioclase peridotites, that melt impregnation, via melt/rock reaction, causes interstitial crystallization of plagioclase and gabbro-norite micro-aggregates, that occurs at temperature conditions lower than solidus conditions of the host peridotite. Plagioclase and gabbro-noritic materials don't crystallize from melt formed by in situ partial melting of the host peridotite, but from exotic melts migrating from the asthenosphere, which infiltrate into and crystallize within the mantle lithosphere. Such liquids modified their composition by melt/peridotite interaction to attempt to reach equilib-rium with the host peridotite, during high melt/rock ratio and open system conditions, towards increasingly silica-saturated derivative liq-uids [55][56][57]. Plagioclase + orthopyroxene impregnation and gabbro-norite pods represent, accordingly, the crys-tallization product of the silica-saturated, strongly trace element deplet-ed basaltic melts. They attained silica-saturation during previous melt/ peridotite interaction under spinel-facies conditions and were not in chemical equilibrium with the host peridotite that they infiltrated by porous flow under shallower plagioclase-facies conditions. More simply, depending on physical and compositional characteris-tics of the system, plagioclase impregnation can occur within the whole plagioclase-facies stability field in mafic and ultramafic systems, when thermal conduction (i.e.,decreasing temperature) induces its crystallization in the pertinent system. Accordingly, plagioclase impregnation does not occur at melting conditions but, more realistic, at subsolidus conditions when melt begins interstitial crystallization of plagioclase and gabbro-norite micro-aggregates, due to prevailing thermal conduction (i.e., conductive cooling) on melt thermal advection (i.e., advective heating). Accordingly, plagioclase crystallization during impregnation is in equilibrium with a derivative melt, modified by melt/rock reaction, and certainly not in equilibrium with a melt formed by in-situ partial melting of the host peridotite. Subsolidus experiments by Borghini et al. [58], performed at pressures lower than 1.0 GPa and temperatures ranging from 900 to 1200°C, on fertile and depleted anhydrous lherzolites, show that a plagioclase-bearing assemblage is stable up to 0.7-0.8 GPa (at T in the range 1000-1100°C), whereas in the depleted lherzolite the upper limit of plagioclase stability is shifted to lower pressure.
Accordingly, for the sake of simplification, we consider the subsolidus (metamorphic) spinel-to plagioclase-facies transition in natural peridotites at P = or b 1.0 GPa, at the temperature condi-tions evaluated in the paper.

The Lithosphere-Asthenosphere Boundary
In our schematic and oversimplified figures we simply report the contact between lithosphere and asthenosphere without any interpretation of the boundary. As reported by Green et al. [59] the reason for differentiation at the lithosphere-as-thenosphere boundary (LAB) is currently being debated with relevant observations from geophysics (including seismology) and geochemistry (including experimental petrology). Water is thought to have an important effect on mantle rheology. Anderson [60] sustained that "Lithosphere is a mechanical concept implying strength and relative permanence. Unfortunately, the term has also been applied to the sur-face thermal boundary layer (TBL) and a shallow enriched geochemical reservoir, features having nothing to do with strength". As sustained by Fischer et al. [61] seismological models provide new constraints on the physical and chemical properties that differentiate the lithosphere from the asthenosphere. The oceanic lithosphere may correspond to a dry, chemically depleted layer over a hydrated, fertile asthenosphere. In this case, observed seismic velocity gradients at LAB require a contrast in mantle hydration, fer-tility, and/or melt content, perhaps in combination with a vertical gradient in velocity anisotropy. Beneath cratons, some studies con-clude that the cratonic lithosphere-asthenosphere boundary is grad-ual enough to be matched by a purely thermal gradient, whereas others indicate a more rapid transition and a contrast in composition or perhaps melt content. Panza et al. [62] and Doglioni et al. [63] outlined the existence of viscosity contrast and decoupling between lithosphere and asthenosphere, and suggested that theboundary is a shear zone. Since the discussion of the nature of LAB is out of the scope of our work, we leave the LAB in our figures without any interpretation.
The lithosphere-asthenosphere system in Italy and surroundings has been thoroughly discussed by Panza et al. [64] and Tuamanian et al. Concerning the reference depth of LAB preceding onset of pas-sive extension, the recent work of Brandmayr et al. [65], "obtained picture of the lithosphere-asthenosphere system for the Italic region which confirms a mantle extremely vertically stratified and laterally strongly heterogeneous. The lateral variability in the mantle is interpreted in terms of subduction zones, slab dehydration, inherited mantle chemical anisotropies, asthenospheric upwellings, and so on", i.e., the recent of actual geodynamic processes acting in the region. In the figures, we use an arbitrary, maximum reference depth of 130 km, taking into account that our data suggest that the lower lith-osphere should have been under garnet-peridotite facies conditions (P N 2.5-3.0 GPa).

the lithosphere
Ongoing field, structural and petrologic studies reveal the presence, within the lherzolite shear zones of the Erro-Tobbio (Voltri Massif) peridotites, of decimetric-metric parallel bands which show peculiar structural and compositional features, with respect to the host lherzolite shear zone. In fact, they show recrystallized granular textures, which replace almost completely the strongly deformed structures of the previous shear zone, reactive micro-structures (pyroxene dissolvingolivine precipi-tating), strong pyroxenes depletion and harzburgite composition.
As discussed by Piccardo (in press), the structural, micro-textural and compositional features of these harzburgite bands, similar to the km-scale bodies of pyroxene-depleted spinel harzburgites of the Erro-Tobbio Unit, are interpreted as records of melt/peridotite inter-action processes during diffuse reactive porous flow percolation [66].
Accordingly, the reactive harzburgite bands within the shear zones indicate that, after onset of asthenosphere partial melting, asthenospheric melts infiltrated along the pre-existing extensional shear zones and percolated by focused and reactive porous flow.
Recently, Mohajeri et al. [67] reported that, while the porosity observed in their simulations is probably not high enough to develop the magmatic dominated final stages of rifting and continental breakup, it would further aid in weakening the lithosphere and promoting localiza-tion of deformation and melt flow. Recent conceptual models based on experiments and field observations allow to explain the schematic rifting stages outlined above [68,69]. These processes could in turn enhance continental thinning and melt supply along the rift axis and eventually [70] could result in continental breakup and the forma-tion of a spreading center (where extension is predominantly accom-modated by magma accretion).

Mohajeri et al. stated that this outcome provides insights into the formation of melt features and shear structures observed in extensional regimes.
Accordingly, the structural and rheological features of the patterned and structured lower lithosphere (i.e. the presence of melt-free porosity bands and shear zones) of the Ligurian peridotites allowed and enhanced melt infiltration via diffuse and focused, reactive porous flow through the host peridotite and melt uprising to shallower levels. Melt saturation and crystallization under plagioclase-peridotite facies conditions. Melt stagnation and storage in the shallow lithosphere evidenced that liquids rising adiabatically from the asthenosphere dissolve pyroxenes and crystallize olivine, that is induced by rapid reaction and slow cooling, during which liquid mass increases. According to Kelemen et al., the melt/peridotite reaction process during upwards melt porous flow mi-gration gradually depletes in pyroxenes the percolated peridotites forming reactive spinel harzburgites, whereas the percolating melts attained silica(− orthopyroxene)-saturation, and were transformed into silica(− orthopyroxene)-saturated derivative liquids, formed by melt-peridotite reaction.
Structural and petrologic investigations on the Ligurian ophiolitic pe-ridotites indicate that these silica (−orthopyroxene)-saturated, deriva-tive liquids, when migrated to relatively shallow mantle levels (i.e., plagioclase-peridotite facies conditions), reacted with the host perido-tites, dissolving olivine and crystallizing orthopyroxene + plagioclase. When they reached their liquidus temperature under plagioclase-peridotite facies conditions, they caused extensive impregnation and refertilization of the shallow lithospheric mantle.
These derivative melts pre-served the strongly depleted geochemical signature of their primary melts (i.e., strong trace element depletion attained by low degrees of fractional melting of a DM spinel-facies mantle source).
Mantle peridotites, that were exhumed and exposed at the seafloor at more distal settings during Jurassic continental breakup in the Liguri-an Tethys domain, consist of plagioclase-bearing/enriched peridotites. The presence of hot melt-impregnated shallow lithosphere indicates the relevance of processes of melt thermal advection that progressively modified the geothermal gradient towards a ridge-type geotherm. Time constraints to mantle evolution from continental rifting to oceanic spreading.
The tectonic and magmatic records evidence the mantle processes that were active in the lithosphere-asthenosphere system during passive extension of the lithosphere to oceanic sea-floor spreading: Shear zones within (ex-garnet-)pyroxenites from the External Ligurides yielded Lu-Hf isochrons which indi-cate a minimum age of 220 +/− 13 Ma for subsolidus transition to plagioclase-facies assemblages. Amphibole-bearing shear zones from the Malenco peridotites [71][72][73] yielded 40Ar/39Ar ages (225 Ma) on amphiboles. These data indicate Upper Triassic subsolidus, melt-absent mantle processes during ongoing extension and exhumation of the mantle lithosphere and suggest that a significant exhumation of the lithospheric mantle has been already active during Upper Triassic, driving lower mantle lithosphere to shallower levels. Accordingly, these Upper Triassic ages cannot indicate onset of passive rifting but may represent intermediate steps of the a-magmatic passive rift evolution, which should have been started significantly earlier, during Triassic (or even during transition from Permian to Triassic).
These few data seem to indicate that melt percolation through the mantle lithosphere, related to asthenosphere partial melting, covers a broad range 175 Ma-155 Ma (from Middle to Upper Jurassic), which apparently is relatively older in the marginal subcontinental peridotites and younger in the distal oceanic peridotites. Ocean opening and MORB extrusion. Bill et al. [85] revised the biochronology of supra-ophiolitic radiolarites in the Alps and Apennines and provided information on the onset of oceanic spreading in the Alpine Tethys and of oceanic MORB extrusion, since basaltic lava flows are frequently interbedded with the radiolarites. They suggested that biochronologic and isotopic ages currently indicate that oceanic spread-ing of the Alpine Tethys began during the Bajocian (Middle Jurassic) and continued until the Kimmeridgian (Upper Jurassic).
Accordingly, it can be extrapolated that onset of oceanization of the Ligurian Tethys basin and MORB eruption occurred during Middle to Upper Jurassic.
The presence/abundance of plagioclase peridotites in modern slow-ultraslow spreading oceans Von der Handt et al. [86] showed the increasing importance of re-active melt stagnation at upper mantle and lower crustal conditions with decreasing spreading rates, at slow-to ultraslow spreading ridges. They sustained the importance of melt-rock interaction at decreasing melt mass between refractory upper mantle rocks and primitive or evolved melts.
Processes of melt/peridotite interaction in abyssal peridotites, lead-ing to formation of plagioclase peridotites, have been recognized and described since a long time samples from modern slow-ultraslow spreading oceans. Plagioclase peridotites, related to melt percolation and impregnation, have been found in MAR, Gakkel Ridge and SWIR [87][88][89][90][91][92][93][94][95][96][97][98]. Dick discussed the presence and local abundance of plagioclase peridotites at SW Indian and America-Antarctic ultra-slow spreading ridges and put in relation the locally-abundant plagio-clase peridotites, where the plagioclase crystallized from impregnated trapped melt (up to 30%), with non-uniform melt flow uprising beneath ocean ridges. According to von der Handt et al., few studies have been devoted to abyssal plagioclase peridotites, despite their relatively high abundance (30% of abyssal peridotites), probably because intense alter-ation obliterates textural relationships and limits the spatial control of the geochemical studies. Although widespread alteration makes it diffi-cult to recognize structural and chemical records of these processes, percolation and crystallization of highly depleted melts in mantle sequences from slow-ultraslow spreading settings have been recognized on the basis of the compositions of clinopyroxenes from peridotites and websterite layers. They sustained that this process produced variable extents of meltrock reaction, dunite formation, and melt impregnation. More-over they found records of cryptic metasomatism by small volumes of late transient silica-rich melts infiltrating through the shallow mantle.
Direct evidence for such melts is seen in orthopyroxenite veins. Dick et al. report that plagioclase peridotites are generally abundant in transform fault zones and at a few localities, such as the Romanche FZ, contain large volumes of plagioclase (2-18 vol.%).
Paganelli et al. described spinel and plagioclase peridotites dredged in the Southern Ridge Transform Intersection (RTI) (SWIR) of the Andrew Bain Fracture Zone (ABFZ). The overall textures of these samples account for important melt percolation/stagnation events occurred in the plagioclase and spinel field. They discussed the causes of inhibition of melt segregation during melt/rock interaction and melt ac-cumulation at depth, i.e. the formation of short scale permeability bar-riers beneath ABFZ or, alternatively, an anomalously thick conductive layer.
Peridotite samples were dredged at the eastern SWIR, between the Melville fracture zone and 63°67′E, smooth seafloor and constitutes about 50% of the mapped area. The collected sam-ples are dominated by spinel peridotites but plagioclase-bearing perido-tites represent about 30% of the mantle derived samples in dredges Dr21 and Dr23 and the near totality of those of Dr14. Seyler et al. selected 24 samples of spinel peridotites, devoid of magmatic veins, lacking plagioclase or its alteration products and containing low-Ti spinels (TiO 2 b 0.2%), to investigate local to regional trace element heterogeneities, and gain some insight into the length scale of the melting heterogeneities and as-sess the respective effects of melting-induced and source inherited heterogeneities.

Snow et al. presented new observations on Lena Trough in the
Arctic Ocean that bear on the early phase of oceanic spreading in such rifts. They described that, at the center of Lena Trough, where no basalt is present, a greater proportion of veined and plagioclase-bearing peridotites occur, which are considered records of melt impreg-nation features related to non-eruptive, syn-rift magmatism. Litho-spheric entrapment of melts has thus been more efficient in the center of Lena Trough, producing thick lithosphere and thin (or non-existent) eruptive crust.
In the following, we concentrate on oceanic plagioclase peridotites (and a gabbro-norite body) which have unaltered clinopyroxene and plagioclase, showing major and trace element equilibration, to unravel some compositional characteristics of the percolated parental melts. The empirical thermometers of Seitz et al. [99] have been applied to the Gakkel Ridge plagioclase peridotites. They furnish very high TSc (T 1224-1237°C), which suggests important melt thermal advec-tion processes to shallow lithospheric mantle levels at the Gakkel ultra-slow spreading ridge.
Few unaltered plagioclases have been found in peridotite samples drilled in the Atlantis II Fracture Zone on the Southwest Indian Ridge. Clinopyroxene rims showing Eu anomaly (i.e., trace element equilibrated with plagioclase) have low Na 2 O con-tents (in the range 0.18-0.39 wt.%), low Sr contents (in the range 0.80-6.17 ppm) and low Zr contents (in the range 1.61-2.88 ppm), Mg# ranges between 90.8 and 92.2. Plagioclases are highly anorthitic (An in the range 83.6-93.2) and have low Sr (in the range 1.4-57.8 ppm) and low Zr (in the range 0.12-0.46 ppm) contents, relatively to plagioclase in equi-librium with aggregated MORB. The strongly trace element depleted minerals compositions suggest a strongly depleted composition of the parental melt.
Peculiar mafic-ultramafic gabbro-norite cumulates, composed by plagioclase lherzolites, olivine gabbro-norites, gabbro-norites and noritic anorthosites, were sampled at DSDP Site 334 on the Mid Atlantic Ridge [100,101]. They contain olivine (Fo 85-90), high-Ca and low-Ca pyroxenes (Mg# 70-91) and plagioclase (An [75][76][77][78][79][80][81][82][83][84][85][86][87][88][89][90] and orthopyroxene is an abundant cumulus phase. Plagioclase has anomalously high anorthite content and pyrox-enes have very high Mg# and Cr abundances, coupled with low Na and Ti in high-Ca pyroxenes and very low incompatible trace element abundances in pyroxenes. Abundance and early crystallization of magnesian orthopyroxene suggest that parental magmas of Site 334 cu-mulates were high silica (52-55 wt.%) liquids. According to Ross and Elthon, the mineral compositions from these oceanic cumulates indicate that the rocks crystallized from basaltic liquids that were strongly depleted in Na, Ti, Zr, Y, Sr and Rare Earth Elements relative to any erupted MORB, that have been collected at crustal levels and remained sufficiently isolated to form distinctive cumulates. Ross and Elthon showed that fractional melting of the upwelling sub-oceanic mantle produces magmas with a much wider range of compo-sitions than erupted MORBs and it seems that strongly depleted primary magmas are routinely produced by melting beneath ridges [102]. These authors noted the absence of similar strongly depleted melts as erupted lavas.

Conclusions
This paper presents an updated review of the field, structural and petrologic constraints recorded by the mantle peridotites from ophiolite massifs that were induced by passive rifting leading to seafloor spreading.
Continental extension induced by passive rifting in the pre-Triassic Europe-Adria domain, caused lithosphere stretching and thinning by melt-free extensional shear zones and consequent almost adiabatic passive upwelling of the shallow asthenosphere (a-magmatic passive rifting). When lithosphere was thinned to half of his thickness, the passively upwelling asthenosphere reached its melting conditions on decompression under spinel-peridotite facies conditions. Reservoir rocks may be developed in pre-rift, syn-rift and post-rift sequences. Effective regional seals may be present within the post-rift sequence if mudstones or evaporites are deposited. Just over half of estimated oil reserves are found associated with rifts containing marine syn-rift and post-rift sequences, just under a quarter in rifts with a non-marine synrift and post-rift, and an eighth in non-marine syn-rift with a marine post-rift. Source rocks are often developed within the sediments filling the active rift (syn-rift), forming either in a lacustrine environment or in a restricted marine environment, although not all rifts contain such sequences.